Fox ESS H1-5.0-E ~33KW Peugeot E208

[Kevin]

Hi,
I seem to have a problem the inverter never feed omre than 37w to the load even thou it is reading 700w, I thought it would try and match the load from the batteries so the grid current is roughly 0?

Here is the BMS data

And the results from your excellent spreadsheet

And what the inverter is reporting the battery as

The inverter is set to feed in first, out max power, 0 export to grid but also tried upping this and made no change
And suggestions please

[woleg]

What is your grid load like? Are you drawing from the grid? Where is your CT clamp?

[Kevin]

My grid load is the garage that the inverter is installed, so has about 400w of florescent tube lighting, a computer and a few other chargers etc, about 700w, the CT clamp is on the incoming mains to the garage, with as per instructions the arrow on the CT clamp pointing towards the grid, so it's simulating the garage is the house , just while I get it all set up

[woleg]

Drop your remaining_kWh, * 0.1 should suffice. I assume you don't have 95.7kWh installed.

Set pack_volts_min to 860 (86.0V) and see what happens. I think the minimum spec is 85V (?)

Other than that I would double check your CT setup. If that reading is wrong the inverter won't do anything. It should be on CT1 and it should rapidly monitor the actual load on that circuit.

[Kevin]

hi Woleg,
Thankyou for your suggestions, in trying to get the BMS working, I did a factory reset the other day and forgot to re-enable the CT clamp, doh.
As far as I can tell it is working, supplying the garage from the batteries so 0w from the grid are now being consumed.
Thank you for taking the time to reply and helping me, if there is anything I can do or test please let me know and I will do what I can to help.

[woleg]

One thing you can do is stop sending can frames to the inverter when your cell voltages are out of spec

I'm quite anal about safety, it's all too easy to make a mistake in code. Do you have a BMS monitoring the cells?

[Kevin]

Thanks for this info, I'm sure it will be useful, I hope to include on the software I have written so far in platform io. Here are some pictures of my BMS unit so far, still a working progress.
The cell voltages
IMG_20230303_091006047_HDR.jpg
Cell temperatures
IMG_20230305_125305636.jpg
And the self test of each BMS
IMG_20230303_091048112_HDR.jpg
And the result
IMG_20230303_091034601_HDR.jpg
The graphics needs more aligning etc , but it gives u an idea of what I'm building

Hi, yes this is the BMS monitoring the cell voltages from the original MG ZS cell packs BMS mounted onto of each board, and also running the canbus codeI have upper and lower limits of the cells and a main contactor that disconnects if any of the parameters are outside of what I have specified.

[Kevin]

Sorry as previous post would include photos, see on page 1 is what I am talking about

[woleg]

Excellent. Rather than have a janky BluePill clone consider one of these. They were made for this job

https://github.com/EliasKotlyar/Canfilter
https://www.aliexpress.com/item/1005003112723581.html (random one picked, make sure you get the blue one and flash with an ST-Link v2)

[Kevin]

Ok thanks, it's actually a stm32f767 using LVGL for the graphics, my own PCB design a 7" touch screen and my code using platform io,

[woleg]

Oh! That's pretty cool. I'll shut up on that front then

[Kevin]

Hi, have been doing some testing and looking through your canbus data that is included with the spreadsheet, and found that if I send the 1881 & 1882 message the inverter only requests this once and once it has received the data it populates the battery serial number in the menu and then starts sending the time stamp as well.

[woleg]

I think that cell voltage data goes in there too but I don't have a V2 FoxESS BMS master to get a can log from. You can see cell voltages on their cloud platform, numbered 1-141 IIRC.

[Kevin]

hi,

I have left inverter running for 24hrs now, just on batteries alone and all is working well including charging at different current rates depending on cell volts. Here is a snippet of what the BMS has received and sent.
Inverter send 13:07:48:053 Data frame Extended frame 00001871 8 02 00 01 00 01 00 00 00 assuming Slave BMS request
Inverter send 13:07:48:563 Data frame Extended frame 00001871 8 03 06 17 04 05 0d 07 2f time stamp
Inverter send 13:07:49:074 Data frame Extended frame 00001871 8 01 00 01 00 02 00 00 00 ?
Inverter send 13:07:49:087 Data frame Extended frame 00001871 8 01 00 01 00 04 00 00 00 ?
Inverter send 13:07:49:583 Data frame Extended frame 00001871 8 01 00 01 00 00 00 00 00 message BMS replying to
BMS reply 13:07:49:844 Data frame Extended frame 00001877 8 00 00 00 00 53 00 1d 20 Bms ASCII data
BMS reply 13:07:49:866 Data frame Extended frame 00001872 8 f0 03 e8 02 32 00 fa 00
BMS reply 13:07:49:876 Data frame Extended frame 00001873 8 6a 03 ec ff 28 00 28 00
BMS reply 13:07:49:893 Data frame Extended frame 00001874 8 c2 01 32 00 2a 00 1f 00
BMS reply 13:07:49:906 Data frame Extended frame 00001875 8 60 00 02 00 01 00 00 00
BMS reply 13:07:49:917 Data frame Extended frame 00001876 8 01 00 3d 0e 00 00 36 0e
BMS reply 13:07:49:927 Data frame Extended frame 00001878 8 6a 03 00 00 98 26 00 00
BMS reply 13:07:49:937 Data frame Extended frame 00001879 8 00 08 01 02 01 02 00 03

Does anybody know what the are? could it possibly be individual/group battery cell voltage info?
Inverter send 13:07:49:074 Data frame Extended frame 00001871 8 01 00 01 00 02 00 00 00 ?
Inverter send 13:07:49:087 Data frame Extended frame 00001871 8 01 00 01 00 04 00 00 00 ?

anything i could try? like i say all is working but just wondering.

[woleg]

I think the other 0x1871 frames are for can slaves, I noticed differences on the BMS version screen when responding to them. Can’t be sure though.

[EV_Builder]

Edit: Nevermind! Found it. Good job en progress

[usus]

Hi,

I haven't posted for a long time but in the last few weeks I restarted the project.
Now is around ~77 KW in 2 packs ~350 Max V.

[usus]

Update

[Kevin]

hi,
Its been a while thought u gave up on the project, glad to see u havent and wow that is alot of storage
Have u had the batteries and inverter running together yet?
my setup is still working, a few minor changes but so far have produce 1800+ KWH from solar and using octopus agile tariff. I set the batteries to charge when the price is low and when solar isn`t suppling enough energy to run the house. Getting more common now as the daylight and sunlight is reducing, but still think i`m saving alot as sometime they pay me to charge the batteries. So far this project has saved me about £600 i think that`s pretty good in the 3 months i`ve had it running with the solar arrays. And an inital outlay of 6K should get my money back pretty quickly.

[usus]

Hi,
Thanks for asking.
I don’t have an actual photo but last that i have in my phone is before connecting to inverter.

Now everything is connected.
I can run but basically, but I am still in safety testing.
I finish testing upper limits and I am discharging slowly for lower limit.
Still balancing is not running. Pack unbalance is ~40 mv.

I don't have solar so for me most saving will be in the Winter months.
Quite hard to quantify the savings, because in Sweden the energy price is variable, but I expect round 60%.
Next step will be an automation for charging on lowest price. https://www.nordpoolgroup.com/en/Market ... view=chart

[usus]

Hi guys!
Sorry to inform you but, look like a kill my FoxEss inverter or at least the CAN communication.
When I start the inverter is sending 1871 but when i try to send any response 1871 is not send any more the 1871 and the inverter doesn’t receive data even if I manually send them.

If anyone has any idea of fault or if can be repair let me know.

The issue appears when I tried to connect another invertor on the back-up AC plug of FoxEss.

BR.

[Kevin]

Oh dear that's not good, I have the backup plug running some stuff in my garage, like 3d printer, a network switch and a rpi4 that I'm using home assistant on to locally control the inverter. Have they done a software update?

[Kevin]

Just a thought, wonder if there is a fault with the backup as this will only run with batteries connected, so it is possible that when it sees a battery it is trying to enable the backup , so not sure if u have but it could be worth turning off the backup supply in the software config and see what happens?

[usus]

hi,
thanks for replay Kevin.
I actually very happy now.
I burn 2 MCP2515 interfaces. one was connected to Foxess inverter the other one to a Sungrow inverter both being connected to the same laptop. strange enough Arduino’s are ok.

@Kevin I don’t see any issue with your 3d printer, a network switch and rpi4.
If they use switching power supply you actually can power them directly from ut to 400V DC. They start with a bridge rectifier that convert the AC in DC. The polarity doesn’t meter.

[Kevin]

Hi, that is good news , I think what happened u had a ground loop between the inverters and that's what blew the can bus ic, , I am using isolated can bus just in case that happened to me. They are more expensive but think worth the extra.
Funny u should suggest about powering directly of the DC of the batteries I am I the process of using an isolated switching converter from HV DC to 12dc for the pi and network switch, thanks for the info thou